Barley (Hordeum vulgare) ranks in fourth place among cultivated cereals for worldwide production and is a recognized model organism for genetic and genomic studies in the Triticeae tribe, which includes wheats (Triticum species) and rye (Secale cereale). Root and shoot architecture traits are key factors in plant performance, competition with weeds, adaptation and stress responses thus having an important impact on yield and yield stability. Breeders have proposed hypothetical optimal morphological parameters to improve production in relation to different environmental conditions. Leaf size and orientation are determinants of canopy transpiration and radiation interception e.g. in dry and sunny Mediterranean environments reduced size and erect orientation of the leaves can reduce water loss by transpiration and allow deeper light penetration into the canopy. Tillering influences crop performance, biomass and grain production, e.g. a reduction in tillering compensated by an increase dimension and number of kernels per spike could be a strategy of adaptation to dry climates. A reduction in plant height and an augment in stem thickness is connected to lodging resistance. Root system extension is connected to the ability of the plant to reach water. The objectives of this project were to dissect genetic variability for shoot and root morphological traits in barley, identifying genomic regions and characterizing genes controlling these traits, and exploring how different traits influence each other. To this end, two approaches were undertaken depending on the trait(s) under study: • the first exploited natural variation in a panel of modern and old European barley cultivars to carry out association mapping of flowering date, stem diameter, spike fertility, leaf dimension, plant height, tillering and root extension (Chapters 2 and 3); • the second was to characterize the ontogenetic basis of increased tillering using as a case study the many-noded dwarf6.6 (mnd6.6) high tillering barley mutant (Chapter 4). In the first approach, we focused on winter barley because of its agronomic interest in the Mediterranean area, where genetic improvement of drought tolerance is particularly important. We analyzed a panel of 142 European winter barley cultivars (67 two-rowed and 75 six-rowed) with a view to conduct a genome wide association scan (GWAS) for shoot and root architecture traits in two separate sets of experiments. To this end, genotyping data for 4,083 SNPs were available from previous projects of which 2,521 mapped on the POPSEQ barley reference map. PCoA results indicated the existence of two major sub-populations in our germplasm panel, corresponding to two-rowed and six-rowed barley cultivars. In order to study shoot developmental traits (Chapter 2) the panel was phenotyped during the growing season 2012-2013 in a field trial at Fiorenzuola d’Arda, Piacenza, Italy. The experimental scheme consisted in 3 replicates (each being a plot of 24 well spaced plants) in randomized blocks. For selected traits data were integrated and analyzed together with those coming from a parallel field trial that was carried out at the University of Shiraz, Iran (data courtesy of Dr. Elahe Tavakol). Flowering date (FD) and leaf width (LW) were measured in both Italy and Iran, leaf length (LL) was measured only in Iran, plant height (PH), spike length (SL), number of fertile rachis node per spike (NFRN) tillering (T) and (SD) were measured only in Italy. Best Linear Unbiased Estimators (BLUEs) of FD, LW were calculated as the phenotypic values estimated for each genotype in a mixed linear model, where genotypes were set as fixed factor and location, location-genotype interaction and replicates as random factors. For BLUEs calculation of all other traits only replicates were used as random factors. BLUEs were subjected to GWAS analyses, using a mixed linear model (MLM) correcting for population structure with a Q matrix (PCA first three coordinate) and for individuals co-ancestry using a K matrix (a pair-wise matrix defining the degree of genetic covariance among individuals). Significance of marker-trait associations was evaluated based on false discovery rate (FDR)-adjusted p-values (threshold value for significant association was set at 0.05). All traits except tillering exhibited good heritability. Few QTLs were detected in GWAS (five for FD, two for LW, three for LL, one for PH, two for SL, two for NFRN, one for tillering, no one for SM). Flowering date exhibited significant correlation with leaf dimension and spike length and six markers designed on Photoperiod-H1 (Ppd-H1) gene (the major determinant for photoperiod response in barley) were the most significantly associated to FD, LW, LL and SL. In particular the recessive ppd- H1 allele causing reduced photoperiod sensitivity, delayed flowering date and increased leaf dimension and spike length compared to the Ppd-H1 allele. Three markers diagnostic for the HvCEN gene (which regulates flowering date independently from photoperiod) were significantly associated to FD and SL. These results suggested that genes for flowering date could have pleiotropic effects on other morphological traits that may mask other genetic effects. For this reason we tested a novel approach repeating GWAS for LW, LL and SL using flowering date as a cofactor (fixed effect) in further analyses. For SL and LW no new significant associations were found with this method, while new significant associations were uncovered for LL, including two markers on chromosome 5H mapped in a region where narrow leaf dwarf 1a (nld-1a) mutant had previously been previously mapped. Tillering and NFRN were only associated to markers diagnostic INTERMEDIUM-C (INT-C), one of the two main genes controlling row type: in our panel two-rowed genotypes had a significantly higher number of tillers and NFRN compared to six- rowed varieties, confirming the known pleiotropic effects of row-type genes on tillering and NFRN and the balancing of patterns of development by breeding practice for the particular row-type. Based on these results, we run GWAS for NFRN and tillering using row-type as covariate. With this model, we found six markers associated with NFRN on chromosome 5H, in the region hosting HvCO12, HvCO13, HvCO15, XvCCA-1, HvLHY, genes involved in control of flowering date. These same markers, were associated to the duration of the phase between awn primordia formation and tipping (awn arising from flag leaf) in a recently published GWAS study. Together, results from Chapter 2 provide the first evidence of the involvement of the Ppd-H1 gene in control of leaf size and spike length. Thus few QTLs were detected that explain the phenotypic variation for our morphological traits, with some major genes having strong pleiotropic effects that mask minor genetic effects. The use of traits that appear to influence others measures as covariates in GWAS models seems to be a promising approach, although the statistical power of this strategy is still to be evaluated. Germplasm collections with uniform growth habit and row-type are an attractive alternative to prevent confounding effects and allow additional loci to be detected. In Chapter 3, we explored natural genetic variation in root extension using the same winter barley panel as Chapter 2 in growth chamber experiments. In order to evaluate root growth we built 50 cm deep cylindrical pots (called rhizotrons) and used digital scans of the root system to measure total root extension with the winRHIZO software. Based on a series of preliminary tests, we used siliceous sand supplemented with controlled release fertilizer to analyze 4th leaf stage plants from 31 genotypes (9 plants per genotype). Root extension per se exhibited 75% heritability, while normalizing root extension on shoot dry weight resulted in low variability (22%) likely due to low heritability of shoot dry weight in our system. These results support the validity of our protocol for evaluation of genetic variation in root extension in barley and other cereals and indicate significant variation exists in our germplasm panel. Thus, the already collected material will be analysed to phenotype the entire panel. In the future, more variability may be uncovered by exploring wild barleys (Hordeum vulgare spp. spontaneum) or landraces. Tillering is a plastic trait affected by the complex interplay of genetic and hormonal factors with environmental conditions such as plant density/light quality and nutrient availability, which likely complicated genetic dissection of this trait in our field experiment on the winter barley panel (Chapter 2). To circumvent the limited power of the GWAS approach for this trait and understand more about the mechanisms subtending tiller formation, we decided to use the mnd6.6 mutant as a case study to investigate the ontogenetic basis of high tillering in barley and its relation to leaf development. Mutant and wild-type plants were grown in growth in a controlled chamber under long day conditions, and dissected weekly from the emergence to anthesis, registering the development of axillary buds, leaves and tillers together with internode elongation, in relation to shoot apical meristem (SAM) stage. Results show that the mutant is not altered in timing of apical meristem development and differentiation to spike, but has a shorter phyllochron that leads to an increment in the number of leaves per vegetative axis. This in turn results in a higher number of axillary buds and a higher number of tillers. The HvMND6 gene was recently identified and our results are consistent with the activity of the previously characterized rice homologue PLASTOCHRON1, indicating an evolutionarily conserved link between plastochron/phyllochron duration and tillering. Concluding, while significant genetic variation was identified for various traits within the gene pool of our winter barley collection, variability of morphological traits as leaf dimension was subordinated to the length of vegetative period. Indeed, flowering date is one of the major factors on which breeding practice has worked to adapt barley to different environments. Beyond modern European varieties, barley breeding for new ideotypes should explore wider genetic resources as Hordeum spp. spontaneum or landraces. In any case, the existence of correlations between different phenotypes calls for careful evaluation of sources of traits to avoid undesired effects on other traits, e.g. due to the relation between tillering and phyllochrone, breeding for early plant vigour through shortening phyllochron, may have pleiotropic effects and result in increased tillering whose benefits would have to be evaluated.

GENETIC DISSECTION OF DEVELOPMENTAL TRAITS IN BARLEY (HORDEUM VULGARE) / G. Verderio ; tutor: L. Rossini ; cotutors: E. Tavakol, L. Cattivelli ; coordinator: P. Bianco. DIPARTIMENTO DI SCIENZE AGRARIE E AMBIENTALI - PRODUZIONE, TERRITORIO, AGROENERGIA, 2015 Jan 23. 27. ciclo, Anno Accademico 2014. [10.13130/verderio-gabriele_phd2015-01-23].

GENETIC DISSECTION OF DEVELOPMENTAL TRAITS IN BARLEY (HORDEUM VULGARE)

G. Verderio
2015

Abstract

Barley (Hordeum vulgare) ranks in fourth place among cultivated cereals for worldwide production and is a recognized model organism for genetic and genomic studies in the Triticeae tribe, which includes wheats (Triticum species) and rye (Secale cereale). Root and shoot architecture traits are key factors in plant performance, competition with weeds, adaptation and stress responses thus having an important impact on yield and yield stability. Breeders have proposed hypothetical optimal morphological parameters to improve production in relation to different environmental conditions. Leaf size and orientation are determinants of canopy transpiration and radiation interception e.g. in dry and sunny Mediterranean environments reduced size and erect orientation of the leaves can reduce water loss by transpiration and allow deeper light penetration into the canopy. Tillering influences crop performance, biomass and grain production, e.g. a reduction in tillering compensated by an increase dimension and number of kernels per spike could be a strategy of adaptation to dry climates. A reduction in plant height and an augment in stem thickness is connected to lodging resistance. Root system extension is connected to the ability of the plant to reach water. The objectives of this project were to dissect genetic variability for shoot and root morphological traits in barley, identifying genomic regions and characterizing genes controlling these traits, and exploring how different traits influence each other. To this end, two approaches were undertaken depending on the trait(s) under study: • the first exploited natural variation in a panel of modern and old European barley cultivars to carry out association mapping of flowering date, stem diameter, spike fertility, leaf dimension, plant height, tillering and root extension (Chapters 2 and 3); • the second was to characterize the ontogenetic basis of increased tillering using as a case study the many-noded dwarf6.6 (mnd6.6) high tillering barley mutant (Chapter 4). In the first approach, we focused on winter barley because of its agronomic interest in the Mediterranean area, where genetic improvement of drought tolerance is particularly important. We analyzed a panel of 142 European winter barley cultivars (67 two-rowed and 75 six-rowed) with a view to conduct a genome wide association scan (GWAS) for shoot and root architecture traits in two separate sets of experiments. To this end, genotyping data for 4,083 SNPs were available from previous projects of which 2,521 mapped on the POPSEQ barley reference map. PCoA results indicated the existence of two major sub-populations in our germplasm panel, corresponding to two-rowed and six-rowed barley cultivars. In order to study shoot developmental traits (Chapter 2) the panel was phenotyped during the growing season 2012-2013 in a field trial at Fiorenzuola d’Arda, Piacenza, Italy. The experimental scheme consisted in 3 replicates (each being a plot of 24 well spaced plants) in randomized blocks. For selected traits data were integrated and analyzed together with those coming from a parallel field trial that was carried out at the University of Shiraz, Iran (data courtesy of Dr. Elahe Tavakol). Flowering date (FD) and leaf width (LW) were measured in both Italy and Iran, leaf length (LL) was measured only in Iran, plant height (PH), spike length (SL), number of fertile rachis node per spike (NFRN) tillering (T) and (SD) were measured only in Italy. Best Linear Unbiased Estimators (BLUEs) of FD, LW were calculated as the phenotypic values estimated for each genotype in a mixed linear model, where genotypes were set as fixed factor and location, location-genotype interaction and replicates as random factors. For BLUEs calculation of all other traits only replicates were used as random factors. BLUEs were subjected to GWAS analyses, using a mixed linear model (MLM) correcting for population structure with a Q matrix (PCA first three coordinate) and for individuals co-ancestry using a K matrix (a pair-wise matrix defining the degree of genetic covariance among individuals). Significance of marker-trait associations was evaluated based on false discovery rate (FDR)-adjusted p-values (threshold value for significant association was set at 0.05). All traits except tillering exhibited good heritability. Few QTLs were detected in GWAS (five for FD, two for LW, three for LL, one for PH, two for SL, two for NFRN, one for tillering, no one for SM). Flowering date exhibited significant correlation with leaf dimension and spike length and six markers designed on Photoperiod-H1 (Ppd-H1) gene (the major determinant for photoperiod response in barley) were the most significantly associated to FD, LW, LL and SL. In particular the recessive ppd- H1 allele causing reduced photoperiod sensitivity, delayed flowering date and increased leaf dimension and spike length compared to the Ppd-H1 allele. Three markers diagnostic for the HvCEN gene (which regulates flowering date independently from photoperiod) were significantly associated to FD and SL. These results suggested that genes for flowering date could have pleiotropic effects on other morphological traits that may mask other genetic effects. For this reason we tested a novel approach repeating GWAS for LW, LL and SL using flowering date as a cofactor (fixed effect) in further analyses. For SL and LW no new significant associations were found with this method, while new significant associations were uncovered for LL, including two markers on chromosome 5H mapped in a region where narrow leaf dwarf 1a (nld-1a) mutant had previously been previously mapped. Tillering and NFRN were only associated to markers diagnostic INTERMEDIUM-C (INT-C), one of the two main genes controlling row type: in our panel two-rowed genotypes had a significantly higher number of tillers and NFRN compared to six- rowed varieties, confirming the known pleiotropic effects of row-type genes on tillering and NFRN and the balancing of patterns of development by breeding practice for the particular row-type. Based on these results, we run GWAS for NFRN and tillering using row-type as covariate. With this model, we found six markers associated with NFRN on chromosome 5H, in the region hosting HvCO12, HvCO13, HvCO15, XvCCA-1, HvLHY, genes involved in control of flowering date. These same markers, were associated to the duration of the phase between awn primordia formation and tipping (awn arising from flag leaf) in a recently published GWAS study. Together, results from Chapter 2 provide the first evidence of the involvement of the Ppd-H1 gene in control of leaf size and spike length. Thus few QTLs were detected that explain the phenotypic variation for our morphological traits, with some major genes having strong pleiotropic effects that mask minor genetic effects. The use of traits that appear to influence others measures as covariates in GWAS models seems to be a promising approach, although the statistical power of this strategy is still to be evaluated. Germplasm collections with uniform growth habit and row-type are an attractive alternative to prevent confounding effects and allow additional loci to be detected. In Chapter 3, we explored natural genetic variation in root extension using the same winter barley panel as Chapter 2 in growth chamber experiments. In order to evaluate root growth we built 50 cm deep cylindrical pots (called rhizotrons) and used digital scans of the root system to measure total root extension with the winRHIZO software. Based on a series of preliminary tests, we used siliceous sand supplemented with controlled release fertilizer to analyze 4th leaf stage plants from 31 genotypes (9 plants per genotype). Root extension per se exhibited 75% heritability, while normalizing root extension on shoot dry weight resulted in low variability (22%) likely due to low heritability of shoot dry weight in our system. These results support the validity of our protocol for evaluation of genetic variation in root extension in barley and other cereals and indicate significant variation exists in our germplasm panel. Thus, the already collected material will be analysed to phenotype the entire panel. In the future, more variability may be uncovered by exploring wild barleys (Hordeum vulgare spp. spontaneum) or landraces. Tillering is a plastic trait affected by the complex interplay of genetic and hormonal factors with environmental conditions such as plant density/light quality and nutrient availability, which likely complicated genetic dissection of this trait in our field experiment on the winter barley panel (Chapter 2). To circumvent the limited power of the GWAS approach for this trait and understand more about the mechanisms subtending tiller formation, we decided to use the mnd6.6 mutant as a case study to investigate the ontogenetic basis of high tillering in barley and its relation to leaf development. Mutant and wild-type plants were grown in growth in a controlled chamber under long day conditions, and dissected weekly from the emergence to anthesis, registering the development of axillary buds, leaves and tillers together with internode elongation, in relation to shoot apical meristem (SAM) stage. Results show that the mutant is not altered in timing of apical meristem development and differentiation to spike, but has a shorter phyllochron that leads to an increment in the number of leaves per vegetative axis. This in turn results in a higher number of axillary buds and a higher number of tillers. The HvMND6 gene was recently identified and our results are consistent with the activity of the previously characterized rice homologue PLASTOCHRON1, indicating an evolutionarily conserved link between plastochron/phyllochron duration and tillering. Concluding, while significant genetic variation was identified for various traits within the gene pool of our winter barley collection, variability of morphological traits as leaf dimension was subordinated to the length of vegetative period. Indeed, flowering date is one of the major factors on which breeding practice has worked to adapt barley to different environments. Beyond modern European varieties, barley breeding for new ideotypes should explore wider genetic resources as Hordeum spp. spontaneum or landraces. In any case, the existence of correlations between different phenotypes calls for careful evaluation of sources of traits to avoid undesired effects on other traits, e.g. due to the relation between tillering and phyllochrone, breeding for early plant vigour through shortening phyllochron, may have pleiotropic effects and result in increased tillering whose benefits would have to be evaluated.
23-gen-2015
Settore AGR/07 - Genetica Agraria
ROSSINI, LAURA
BIANCO, PIERO ATTILIO
Doctoral Thesis
GENETIC DISSECTION OF DEVELOPMENTAL TRAITS IN BARLEY (HORDEUM VULGARE) / G. Verderio ; tutor: L. Rossini ; cotutors: E. Tavakol, L. Cattivelli ; coordinator: P. Bianco. DIPARTIMENTO DI SCIENZE AGRARIE E AMBIENTALI - PRODUZIONE, TERRITORIO, AGROENERGIA, 2015 Jan 23. 27. ciclo, Anno Accademico 2014. [10.13130/verderio-gabriele_phd2015-01-23].
File in questo prodotto:
File Dimensione Formato  
phd_unimi_R09682.pdf

Open Access dal 29/06/2016

Descrizione: Tesi completa
Tipologia: Tesi di dottorato completa
Dimensione 5.27 MB
Formato Adobe PDF
5.27 MB Adobe PDF Visualizza/Apri
Pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/2434/252988
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? ND
social impact